Review of Peters, Pam and Kate Burridge eds. 2023. Exploring the Ecology of World Englishes in the Twenty-first Century: Language, Society and Culture

This volume is a collection of articles on the relation between language and its context. It derives from a project called Varieties of English in the Indo-Pacific,¹ so the language is English as a first, second, or foreign language and the majority of the chapters deal with English in Australasia or the Pacific islands. While many chapters are based on corpus analyses of varying degrees of sophistication, readers of this journal may not learn much about technical or statistical aspects of corpus study. Instead, the focus of the book is on how linguistic features, or their frequency, can be related to the cultural or sociolinguistic context.

All the chapters are rich sources of information and examples, often about little-discussed forms of English. In the present context readers may be interested in the nature of the sample from which the data are derived. Five are based on established corpora, three on the language of ethnographic and/or sociolinguistic interviews and three on ad-hoc written corpora. Two chapters mainly deal with pronunciation in relation to a multilingual environment, and hence on elicited sample data of various kinds. Two chapters rather stand out, one based on literary texts and the other reinterpreting established knowledge to argue for improved policy.

A framework adopted in many of the chapters is set by Edgar W. Schneider in chapter 2, summarising Schneider (2018) and applying it to Indo-Pacific examples. The idea is that varieties of English reflect local culture in three ways, called ‘nexuses’. The first nexus is the familiar presence in the variety of vocabulary items referring to local culture and nature: Peters and Burridge cite Indian English masala and South African English veldt as examples. The second examines the frequency of characteristic indicator terms, words or phrases (for example we vs. I, sir, must) and views these frequencies as reflecting the types of cultural value parameters examined in the World Values Survey,² in this volume primarily individualism-collectivism (dividing ‘the west from the rest’). The third relates structural schemas (active vs. passive, for example) to values of the same kind. Schneider is appropriately cautious about the framework, but many of the chapters in the volume give support to it, as an explanatory if not a predictive model. Other chapters propose other types of culture-language links or relate linguistic features rather to the sociolinguistics of the user community than to its culture.

Chapter 3, by Bertus van Rooy, examines background Afrikaans in some English-language literary fiction about Afrikaners written by the Afrikaner Herman Charles Bosman. Bosman uses many Afrikaans vocabulary items to give a flavour of the context and culture described (Schneider’s nexuses 1 and partly 2). In terms of frequency, Van Rooy examines specifically phrasal verbs, adverb placement and verb-second word order. He shows, for example, that the Afrikaans calque think out ‘invent’ (rather than think up) is frequent in Bosman. Placement of adverbs between subject and verb and verb-second constructions are somewhat marked and literary in English but obligatory in Afrikaans. Because the examples are all normal standard English, further analysis would be required to determine whether they are marked in Bosman’s prose or appropriate for its stylistic level and period.

Chapter 4, by Loy Liseng, uses the Philippine component of the International Corpus of English (ICE),³ but only the 400,000–word written part to ensure that the words found are reasonably well established in English. Nonetheless, the text-type that is closest to speech ––social letters–– yields the largest number of examples of the target category, which is lexicosemantic tokens from Philippine languages (including Spanish) in Philippine English (PhE). The findings mainly exemplify Schneider’s nexuses 1 and 2 with ‘cultural’ borrowings with uniquely local referents in the expected semantic fields such as costumes, flora and fauna, food, music, units of government and the largest social relationships and kinship terms. There are also several ‘core’ tokens which have equivalents in other varieties but of course local associations, such as merienda ‘snack’.

Chapter 5, by Pam Peters, looks at Indian English (IndE) in the Corpus of Global Web-Based English (GloWbe)⁴ and several older dictionaries to give rich historical depth to a corpus study of frequent IndE loanwords. Words found in GloWbe are traced through dictionaries back to their presence in nineteenth-century IndE and then back to their sources in Sanskrit, Persian or Arabic, mostly via Hindustani. 20 frequent words are identified as ‘keywords’ and their long history in IndE and entwinement with the administrative role of IndE make them revealing instances of nexuses 1 and 2. Crore (‘ten million’) and lakh (‘one hundred thousand’) are the two most frequent examples. These are statistically key words marking IndE, but hardly keys to Indian culture in Raymond Williams’ (2017) sense, so the term ‘keywords’ needs some unpacking.

Chapter 6, by Christiane Meierkord and Bebwa Isingoma, examines greetings, address terms and discourse markers expressing stance, that is nexus 2 words, in Uganda English (UgE). The data are from the written part and the incomplete spoken part of the Uganda component of ICE (ICE-UG), along with a corpus of web-based writings with about 12.3 million words. Although pragmatics shows first-culture influence, only discourse markers show much influence from the linguistic context; there are fewer borrowings from local languages in UgE than from Hindustani in IndE (chapter 4) or Tagalog/Spanish in PhE (chapter 5). Uganda is highly multilingual, so speakers cannot rely on borrowings being understood. English did not become widespread under the protectorate of Uganda as it did in the other two colonised territories. In these circumstances, UgE is more exonormative than other varieties.

Chapter 7, by Sara Lynch, Eva Kuske and Dominique B. Hess, is based on Micronesian English interviews in three locations (Guam, Saipan and Kosrae) which represent different degrees of acculturation to US norms. Guam should be the most westernised with the weakest family values and Kosrae the least westernised with the strongest such values, with the notably multicultural Saipan in the middle. Using words from all three locations, the authors compare some 35 ‘cultural key words’ related to kinship to show that this type of corpus analysis (lexical quantification), is a valid approach to defining culture. No tests of significance were applied, so the results are not easy to interpret, but it is fair to say that the figures do not show a linear increase in kinship terms in general in parallel with the presumed stage of westernisation. Reference must be made to the attitudes and traditions of the individual ethnic groups, and this shows the heuristic advantages of the approach.

Chapter 8, by Hannah Hedegard, also examines spoken English from interviews. Her data come from a very small community, the Cocos (Keeling) Islands, making possible a highly representative sample of a homogeneous group. The island community exhibits a sharp division between those over 50, socialised wholly in Malay, and those under 50, who are fluent users of English and participate in both Cocos Malay and Australian culture. Both the we/I ratio and the frequency of indicator vocabulary indicate that older informants are more collectivist (figures similar to Indian informants) than the younger, who more individualist (even exceeding figures for UK informants in ICE). The incidence of kinship terms is, however, relatively uniform across generations, possibly reflecting the dense multiplex society of the islands. Although the frequency of such terms is described as high, the topics of the current corpus are probably biased in that direction; comparisons must be made across comparable corpora.

Chapter 9, by Kathleen Ahrens and Winnie Huiheng Zeng, examines a 130,841–word corpus of editorials (from carefully-chosen dates, in such a way that there was much discussion of the US 2016 elections) retrieved from English-language newspapers in the ‘Sinosphere’, in this case two based in Beijing, three in Hong Kong and two in Taipei. They investigate the ways conceptual metaphors can express culture; Ahrens and Zeng see this as a fourth ‘nexus’ to add to Schneider’s three. The Hong Kong and Taipei papers represent differing political positions. They search for cultural differences among the metaphors in the domain democracy appearing in the editorials, thus making ideology a subfield of culture. They find a statistically significant tendency for Hong Kong papers to use democracy literally relative to Beijing’s metaphorical uses. Statistically, Hong Kong tends to associate democracy with building (as Consolidating the election platforms), while Taipei with journey (as in a hurdle on Taiwan’s road to democracy). Although Beijing was mostly discussing democracy in the US, Hong Kong and Taipei predominantly discussed local democracy. While the authors call these differences among Englishes, they seem primarily to be differences in content or ideology which might well be found among users of the same English.

In chapter 10, Sarah Buschfeld examines those elements of linguistic variability in the speech of children who are native speakers of Singapore English (SingE), which can be ascribed to the children’s ethnicity. Buschfeld elicited varied speech samples from 30 children and here reports analyses of a few syntactic, morphological and phonological features. The children’s speech was highly variable across and within individuals between BrE and AmE variants; the same child, for example said both /dɑ:ns/ and /dæns/, and another both /ˈpeɪntəd/ and /ˈaɪsskeɪdɪd/. There was similar variation between colloquial realisations of SingE on phonetic and morphosyntactic dimensions. Subject-deletion and past tense realisation show wide variation between British L1 usage and ‘Singlish’, but here there is an ethnic dimension: subject-deletion is more common in the Chinese group than in the Indian one, though both have pro-drop languages in the input. Non realisation of past-tense is also more common among Chinese than Indian children, but, as Indian languages have marked past tenses, what is interesting is that the Indian children have unmarked past, presumably because there is Chinese-influenced speech in the input.

In chapter 11, Tobias Bernaisch and Sandra Götz compare discourse styles in competent-speaker language from the ICE corpora for Great Britain, Hong Kong, the Philippines and Singapore with those in learner language from the International Corpus Network of Asian Learners of English (ICNALE)⁵ for the same three Southeast Asian territories. They ask whether formal differences between discourse styles in English can be ascribed to cultural differences between users or to their acquisition status. Discourse style is operationalised as the relative frequency in the texts of nouns and verbs (independently) and the data are analysed using both conditional inference trees and linear regression – thus a considerably more sophisticated analysis than in some other papers. The results are of course complicated. There is evidence for a specifically Hong Kong ‘nouny’ discourse style across user acquisition types, and for a tendency for spoken and written discourse styles not to be differentiated in ESL usage, while these styles are differentiated in both ENL and EFL usage, implying perhaps greater exonormativity among the learners. These discourse patterns add another possible nexus to Schneider’s nexus three.

In chapter 12, Pam Peters, Tobias Bernaisch and Kathleen Ahrens look at the same corpus as Ahrens and Zeng (chapter 9) and ask whether the use of modals/quasi-modals in newspaper editorials aligns with the newspaper’s political stance and/or reflects the local sociocultural climate. Their analysis of the use of modals is statistically informed, although in corpus analysis of modals the epistemic, deontic or dynamic meanings of the forms are difficult to distinguish. Modals are often deontic and, given the differences in local ideologies, it comes as no surprise that frequencies of modals and semimodals are significantly different between the three territories. These overall differences include cases where there are large differences in the use of a particular form, which predictably align with differences in the stance or function of the English-language press. Beijing editorials use assertive modals such as will and should more frequently than Hong Kong and Taipei editorials, reflecting the tendency for the Beijing English-language press to announce policy rather than to discuss it. But there are also differences related to the local sociolinguistics of English. The Hong Kong editorials are more like colloquial speech and, for example, make a more frequent use of going to, reflecting the continuing second-language status of English.

A related topic is examined diachronically by Adam Smith, Minna Korhonen, Haidee Kotze and Bertus van Rooy in chapter 15, which looks at Australian, New Zealand and British parliamentary records taken from Hansard in the period 1901–2015.⁶ The chapter investigates changes in the use of the (semi-)modal verbs must, have to, need to and should, and the subjects of the verbs (namely, we, the government, the party) are also examined. The changes found are seen as reflecting processes like democratisation, colloquialisation and changes in the imagined audience due to broadcasting. It is shown convincingly that in all three corpora need to has increased considerably since the 1950s. However, there are difficulties with the corpus and the analysis. No account is taken of the editing practices of Hansard (Mollin 2007), which convert have to to must and expand contractions (potentially I’d like to I should like). Likewise, no distinction is made between epistemic (he must know) and deontic uses. Statistically speaking, no tests of significance are carried out, so we do not know which of the very many numerical changes discussed point to real differences.

In chapter 13, Kate Burridge and Carolin Biewer look at published accounts of Australian English and the more ‘exotic’ system of Pitkern-Norf’k. Likewise, they examine Biewer’s collection of interviews in acrolectal South Pacific English (SPE) with 24 Samoans, 24 Fijians and 24 Cook Islands Māori (120,000 words). The study investigates how pronominal systems can relate to local culture and language ecology. The Pitkern-Norf’k language has an elaborate system with singular, dual and plural, exclusive and inclusive first and second person pronouns, both calqued on Tahitian and expressing the strong identity of the tiny community (Schneider’s third nexus, the system is an indicator structure). In acrolectal SPE uses of standard I, we, we all and they are shown in context to reflect local attitudes to community membership and identity (Schneider’s second nexus).

Chapter 14, by Ian Malcolm, uses linguistic analysis and cultural nexuses to argue for changed educational policy. It shows that English has been nativised separately by immigrant and Indigenous communities in Australia, leading to the existence of Australian English (AusE) and Aboriginal English (AbE). Without referring to a corpus, Malcolm provides numerous examples from AbE, showing that its vocabulary, forms of address, and possibly even syntax reflect elements of Indigenous people’s experience, relatable to all three of Schneider’s nexuses. The formation of AbE represents decolonisation of the imposed language, making it the voice of the dispossessed minority. Its exclusion from the educational process is a type of neocolonialism. This can be overcome by aiming at ‘postcolonial biculturism’ rather than at assimilation, within a system which recognises bilingual proficiency in both AusE and AbE as a resource.

In chapter 16, Isabelle Burke and Kate Burridge make use of data from the UWA Corpus of English in Australia (Rodriguez Louro 2022) ––with more than a million words of casual student dialogue–– and discuss linguistic detail. They examine various informal negatives in relation to the ‘Jespersen cycle’ (Jespersen 1917) and focus on the construction I know damn all about it, which they show to be a twentieth-century innovation in which the taboo word + all has always been negative. In Australia, but apparently not elsewhere, expressions such as damn all have progressed from being negative quantifier (‘nothing’) to fully-fledged adverbial negator (‘not’). Colloquial language is an element of the Australian self-image, so that the reanalysis reflects Australian culture, somewhat ironically since it is combined with a strong prescriptive tradition. A piquant example is that their informants strongly rejected I don’t know bugger all about it as ‘double negation’, while accepting the taboo word. The chapter shows a nuanced linguistic awareness not found in the other chapters.

Chapter 17, by Miriam Meyerhoff, Elaine Ballard, Helen Charters, Alexandra Birchfield and Catherine I. Watson looks at the sociolinguistic context of language change. It describes the Auckland Voices Project and tests whether the increasing heterogeneity of urban speech communities suggests new theories about language variation and change. The study examines the spread of the pronunciation of the with a schwa into prevocalic positions. Like Hedegard’s study in chapter 8, the project looks at younger and older speakers in three Auckland communities, one (Titirangi) predominantly Pakeha (European), one (South Auckland) with a long-standing ethnic mix and one (Mount Roskill) in transition from Pakeha predominance to ethnic mixture. The results show that pronunciations of the with shwa before vowels are more frequent in South Auckland than in the other two, and rather more frequent in Mount Roskill than in Titirangi. According to the data, older speakers have fewer shwa pronunciations than younger ones in all three areas, and in all three the quality of the following vowel affects the frequency of schwa. It appears the levelling is, as the writers hypothesised, led by the most diverse community and younger speakers and that it is spreading to speakers in other types of community. This is parallel to the development in London.

Most chapters are valuable for their linguistic examples alone and as introductions to the varieties and projects discussed, and nearly all do indeed cast light on the relation of language forms to their context. They leave the reader with a sense that this relation can be very direct for the multilingualism of the context and for the loanwords in nexus 1. For grammatical words and more abstract structures, there often seem to be more possible explanations of the findings than a connection to the rather nebulous collectivist-individualist cline.